the best of esmo 2016...dm has received honoraria for consultancy/advisory boards and symposia from...

esmo.org

THE BEST OF ESMO 2016

Colorectal cancer

Dr David MALKA

Gustave Roussy

Université Paris Sud

Villejuif, FRANCE

DM has received honoraria for consultancy/advisory boards and symposia from Amgen,

Bayer, Celgene, Lilly, Merck Serono, MSD, Roche, Sanofi-Aventis and Servier, and has

received research funding (to his institution) from BMS and Immune Design

DISCLOSURES

NON-METASTATIC DISEASE

Surgery for rectal cancer: 6-7 or 11-12 weeks

after chemoradiotherapy (CRT)?

W8 W10 W12 W6

TME

Lyon R90-01 Does it increase pCR rate? Sphincter

preservation? Survival?

CRT

6 vs 12 trial

• Primary end point: to increase MRI downstaging from 40% (6w) to 60% (12w)

• Patients: 237 from 21 centres (37% of them included in one centre)

Inclusion criteria?

Bhoday J et al. Abstract 4520. ESMO 2016

• MRI: at initial staging (preCRT) and at 6 +/- 12w

• Then RT then surgery at 6-8w vs12-14w

Is it clinically relevant?

RT or CRT?

6 vs 12 trial


mrT down-staging

52 patients (43%) 6w vs. 67 patients (58%) 12w, p = 0.019

122 patients (6w arm), 115 patients (12w arm)

mrN down-staging

54 patients (44%) 6w vs. 61 patients (53%) 12w, NS

Hypothesis 40 => 60%


6 week

n=122

12 week

n=115

p

mrTRG 1 6% 22% <0.05

CRM involvement 68% 61% NS

EMVI 50% 33% 0.001

ypT0 7.4% 20% <0.05

ypN0 42% 55% <0.05

pTRG 1 = pCR 9% 20% <0.05

• No significant difference between the 2 arms for surgical morbidity.

• 1% anastomotic leak???

Lefèvre J et al. JCO 2016

• 265 patients with T3-T4 and/or N+ rectal cancers from 24 centres

• Randomization after CRT: 7w vs11w interval before surgery


pCR rate (ypT0N0) : 15% (7w) vs 17.4% (11w) p=0.6

+3.6% pCR

Per protocol

ITT


Surgical morbidity 7 weeks 11 weeks

Complete and detailed data needed

6 weeks may be too early

12 weeks may be too late

7-9 weeks seems enough to reach an optimal pCR without morbidity

TO CONCLUDE


NGS ASSESSMENT OF CETUXIMAB ADJUVANT TRIAL

Is there a place?

• Better HR than MOSAIC (0.8)

• Clinically relevant but not significant

• Subgroup analysis (unplanned)

New trial in hyper-selected RAS/BRAFWT patients?

cetuximab

No cetuximab

HR= 0.77

P=0.1

Taieb et al ESM0 2016

Does surveillance matter?

Regular CEA monitoring

Colonoscopy at year 1 and every 3–5 years

Consider CT FU in patients at higher risk

Clinical assessment 6 monthly for 2 years

Hx and colonoscopy every 5 years (up to 75)

Additional FU for patients with symptoms

Regular CEA and CT FU for Stage II-III CRC

Insufficient evidence for stage I disease

Colonoscopy at 1 year and every 5 years

Ann Oncol 2013; 24 (Suppl 6): vi64-vi72.

Ann Oncol 2013; 24 (Suppl 6): vi81-vi88.

J Clin Oncol 31:4465-4470

Does surveillance matter?

301

minimal

299

CT

300

CEA 302

CEA & CT

1211 patients

Pugh S et al. Abstract 4530. ESMO 2016

6-12-year results from the FACS trial

Intensive follow-up identified more recurrences surgically treatable with curative intent

Minimum follow-up

Intensive follow-up

CEA CT CEA + CT P value

n=301 n=300 n=299 n=302

Recurrences (%) 38 (13%) 56 (19%) 61 (20%) 48 (16%) 0.06

Surgically treatable recurrences (%) 8 (3%) 19 (6%) 28 (9%) 21 (7%) 0.008

Among recurrences, treatable (%) 21% 32% 46% 44%

Does it translate in survival?

Pugh S et al. Abstract 4530. ESMO 2016

6-12 year results from the FACS trial

Does surveillance matter? Intention to treat analysis

Log rank p = 0.45

Per protocol analysis

Log rank p = 0.36

• Of course NOT in ITT or PP

• But what is important is

survival in patients that HAVE

RELAPSED!!!

• <100 relapses out of 1200 pts

Yes, surveillance matters, especially CT, as it brings more

patients to curative-intent Tx at relapse!

News from very rare mutations/alterations!

Little is known about rare mutations in stage III colon cancer

Their prognostic value may be important to stratify clinical trials or

intensify adjuvant treatments in the future

They may have predictive value as well!!!

News from very rare mutations/alterations!

Impact of rare individual

mutations on TTR

Favors mutant Favors WT

Rare RAS and BRAF mutations are not

prognostic (except codon 61) POLE is a good prognostic factor

(sensitivity to PD(L)1i?)

Taïeb et al., Laurent-Puig et al., Domingo et al. ESMO 2016

HER2/ERBB2 is a poor prognostic

factor (sensitivity to HER2i?)

10%

4% 1%

METASTATIC DISEASE

What is the best maintenance Tx?

Capecitabine > chemotherapy stop (Luo Ann Oncol 2016)

Bevacizumab alone = marginal benefit (SAKK, PRODIGE 9)

Capecitabine + bevacizumab

largely used

> bevacizumab alone or complete stop (CAIRO3, AIO 0203)

Added value of bevacizumab vs. FP alone possible but still not proven

Maintenance with cetuximab suggested by a R ph II trial (COIN.B)

What is the best maintenance Tx?

MACBETH: Study design

R 1:1

mFOLFOXIRI +

cetuximab§ up to 8 cycles

cetuxim ab§

unt i l PD

bevacizumab §

unt i l PD

mFOLFOXIRI +

cetuximab§

up to 8 cycles mCRC pts:

ü Unresectab le disease

ü Previously unt reat ed

f or m ts disease

ü RAS and BRAF w t *

Arm

A

Arm

B

*cent ral ly assessed: KRAS 12,13,61 w t unt i l Oct 2013, then RAS and BRAF w t §administered b iw eek ly St rat i f icat ion f actor : center

Phase II randomized non-comparat ive t r ial

INDUCTION MAINTENANCE

Cremolini et al. ESMO 2016

What is the best maintenance? Macbeth (mITT population)

RASwt

Arm B (N pts=57)

Arm A (N pts=59)

Events

43

50

Median

11.2 mos

9.3 mos

95% CI

7.2-14.0

7.1-10.8

RASWT patients

Cetux seems> bev after cetux induction

In accordance with COIN B results

Still needs to be confirmed

(insufficient number of patients)

Cap (FP?) + cetux should be compared to

• FP + bev

• FP alone!

RASWT subpopulation in all studies

Tumor sidedness definition

Sigmoid

colon

Splenic

flexure

Cecum

Descending

colon

Rectum*

Hepatic

flexure

Ascending

colon

Rectosigmoid

junction

Transverse colon

Left colorectum

Whose side are you on?

Right colon

?

BRAF Mutants

CIMP+

MSI+

Primary tumor not removed

Older patients

PEAK – OS – 1st line

Beva, bevacizumab; HR, hazard ratio; OS, overall survival; Pmab, panitumumab

100

80

60

40

20

0

Kap

lan-

Mei

er e

stim

ate

0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68

53

54

22

14

51

51

21

12

49

48

18

11

46

44

13

11

44

43

11

7

41

40

10

6

36

35

8

4

33

31

7

3

32

26

6

1

31

21

5

0

25

17

4

17

13

3

13

10

2

6

5

1

3

3

0

2

2

0

0

No. of subjects:

1:

2:

3:

4:

1: Pmab + FOLFOX left

2: Beva + FOLFOX left

3: Pmab + FOLFOX right

4: Beva + FOLFOX right

Median OS (95% CI), months

Pmab +

FOLFOX

Beva +

FOLFOX HR (95% CI)

Left 43.4 (31.6–63.0) 32.0 (26.0–47.4) 0.84 (0.22–3.27)

Right 17.5 (9.1–30.7) 21.0 (6.0–29.0) 0.45 (0.08–2.49)

Censor indicated by vertical bar

Overall survival (months)

Peeters et al. ESMO 2016

PFS – 1st (PRIME and PEAK) and 2nd (181) line

Median PFS (95% CI), months

PRIME (1st line) PEAK (1st line) 181 (2nd line)

Pmab +

FOLFOX

FOLFOX

HR

(95% CI)

Pmab +

FOLFOX

Beva +

FOLFOX

HR

(95% CI)

Pmab +

FOLFIRI

FOLFIRI HR

(95% CI)

Left 12.9

(10.0–14.6)

9.2

(7.6–10.7) 0.72

(0.57–0.90)

14.6

(11.6–17.7)

11.5

(9.3–13.0)

0.65

(0.21–2.00)

8.0

(7.3–9.1)

5.8

(5.2–7.3)

0.88

(0.69–1.12)

Right 7.5

(5.5–10.4)

7.0

(5.4–8.0)

0.80

(0.50–1.26)

8.7

(5.7–10.9)

12.6

(1.8–16.6)

0.84

(0.18–3.79)

4.8

(3.5–7.4)

2.4

(1.8–5.7)

0.75

(0.45–1.27)

Beva, bevacizumab; CI, confidence interval; HR, hazard ratio; PFS, progression-free survival; Pmab, panitumumab


Response rates

Complete response + partial response (%)

PRIME (1st line) PEAK (1st line) 181 (2nd line)

Pmab +

FOLFOX

FOLFOX Pmab +

FOLFOX

Beva +

FOLFOX

Pmab +

FOLFIRI

FOLFIRI

Left 68 53 64 57 50 13

Right 42 35 63 50 13 3

Beva, bevacizumab; Pmab, panitumumab


Whose side are you on? (CALGB all RASWT) 80405: Overall Survival by Biologic

(Right Sided Primary) Arm

N

(Events)

Median

(95% CI)

HR

(95% CI) Adjusted p

Bev 78

(58)

29.2

(22.4-36.9) 1.36

(0.93-1.99) 0.10

Cetux 71

(56)

13.7

(11.3-19.0)

80405: Overall Survival by Biologic

(Left Sided Primary) Arm

N

(Events)

Median

(95% CI)

HR

(95% CI)

Adjusted

p

Bev 152

(119)

32.6

(28.3-36.2) 0.77

(0.59-0.99) 0.04

Cetux 173

(119)

39.3

(32.9-42.9)

Venook et al. ESMO 2016

Whose side are you on? (FIRE.3 and CRYSTAL all RASWT)

Heinemann et al. and Van Cutsem et al. ESMO 2016

Whose side are you on? (FIRE.3 and CRYSTAL all RASWT)

Whose side are you on? Conclusion

• In LEFT-SIDED RASWT • 2+anti-EGFR >2+bev for shrinkage and disease control

• Option: 3+bev (FOLFOXIRI+bev) in fit patients

• In RIGHT-SIDED RASWT

• Poor prognosis pts (BRAFmt in 20-30%!)

• BOTH 2+BEV OR 2+ANTI-EGFR REMAIN POSSIBLE

• 3+bev (FOLFOXIRI+bev) may be the good choice in fit patients

→ Clinical trials still needed

Do we have to change our practice based on unplanned subgroup analyses of

1st (2nd) line, non-strategic trials?

esmo.org

MIR-31-3P IS A PREDICTIVE BIOMARKER OF CETUXIMAB RESPONSE IN FIRE-3 CLINICAL TRIAL P. Laurent-Puig, M.L. Grisoni, V. Heinemann, K. Fontaine, C. Vazart, V. Decaulne, F.

Rousseau, B. Courtieu, F. Liebaert, A. Jung, D. Neureiter, R. Thiébaut, S. Stintzing

Treatment effect on PFS, OS and ORR

• Overall: Hazard/Odds ratios [95% CI] not adjusted

p=0.046

HR=0.57 [0.37 ; 0.87] p=0.003

HR=1.13 [0.66 ; 1.96] p=0.61

HR (97.5%CI)

p=0.048

HR=0.80 [0.58 ; 1.12] p=0.13

HR=1.33 [0.82 ; 2.14] p=0.18

HR (97.5%CI)

Low

n = 229

High

n = 111 OR=1.23 [0.54 ; 2.80] p=0.62

OR=4.13 [1.99 ; 8.61] p=0.0002

p=0.028

Favors cetux Favors beva Favors cetux Favors beva Favors beva

• In miR-31-3p subgroups: Hazard/Odds ratios adjusted on age, number of organs and BRAF status2

MiR-31-3p is predictive of treatement effect on PFS, OS and ORR

OR (95%CI) Favors cetux

OS

HR=0.70 [0.53 ; 0.92] p=0.011

HR=0.71 [0.53 ; 0.93] p=0.014

ITT RAS WT1

mITT

ORR

OR*=1.28 [0.83 ; 1.99] p=0.32

OR*=1.66 [1.07 ; 2.59] p=0.03

PFS

HR=0.93 [0.74 ; 1.17] p=0.54

HR=0.92 [0.73 ; 1.16] p=0.46

1Heinemann et al. Lancet Oncology 2014

*Missing data included as non-responders

Interaction test

Treatment effect on early tumor shrinkage (ETS) at 6 weeks

Favors beva Favors cetux

Interaction test: p=0.029

MiR-31-3p is predictive of treatment effect on early tumor shrinkage cetuximab gives an earlier response than bevacizumab in low miR-31-3p expressors

High n = 111 Low n = 229

Responders:

ETS at 6 weeks < -30 %

Odds ratio [95% CI] adjusted on age, number of organs and BRAF status


beva responders

25%

cetux responders

56%

beva responders

28%

cetux responders

35%

n = 47 n = 51 n = 104 n = 77 Excluding patients with missing response

OR = 4.11 [2.14 ; 7.92] p=0.00002 OR = 1.16 [0.46 ; 2.92] p=0.76

Treatment effect on depth of response (DoR)


Interaction test p=0.0006

MiR-31-3p is predictive of treatment effect on depth of response cetuximab gives an deeper response than bevacizumab in low miR-31-3p expressors

High n = 111 Low n = 229

OR=6.05 [2.77 ; 13.22] p=0.000006 OR=0.84 [0.35 ; 2.02] p=0.70

Responders:

depth of response > -30 %

beva Responders

53%

cetux responders

53%

n = 47 n = 51

beva responders

53%

cetux responders

86%

n = 104 n = 77 Excluding patients with missing response

Odds ratio [95% CI] adjusted on age, number of organs and BRAF status


Bevacizumab n=151 Correlation test: cor=0.12 , p=0.159

Cetuximab n=128 Correlation test: cor=0.28 , p=0.002

Depth of reponse by miR-31-3p expression

Non Responders

20%

Stable disease

-30%

Responders

CUT OFF

Low High

CUT OFF

Low High

MiR-31-3p & sidedness

MiR-31-3p expression groups by sideness

Twofold increase in the percent of patients with right sided tumors who

have high miR-31-3p expression.

OS by miR-31-3p subgroups and sidedness Left High

Left Low

Right High n = 69

n = 191 n = 26 Low

High

Left Right

HR=2.85 [1.19 ; 6.85]

HR=0.52 [0.34 ; 0.79]

Right Low

n = 41 n = 111

n = 229

n = 260 n = 67

miR by treat p=0,09 miR by treat p=0,64

Both miR-31-3p and sidedness are predictive: • Right-High pts: OS benefit with bev • Left-Low pts: OS benefit with cetux

Sidedness by treat p=0,001



miR by treat p=0,046

Interaction

test p-value

_ bevacizumab _ cetuximab

2nd line FOLFIRI +/- regorafenib

O’Neil. et al. - ESMO® 2016 - Abs. 464 PD

R A B D O M I Z E D

N = 240

Screening :

Obtain archival tissue for correlative

studies in all patients, and fresh

biopsy (optional) in subset1

mCRC following 1 prior

oxaliplatin-containing regimen

Primary

endpoint :

PFS

Secondary

endpoints :

ORR, DCR, OS, PK,

safety, and tolerability

FOLFIRI†

+ Placebo

†Day 1–2, Day 15–16

Regorafenib 160 mg*

+ FOLFIRI

*Day 4–10, Day 18–24

R A B D O M I Z E D

2 to 1

phase I - Schulteis et al. Ann Oncol 2013

Repeat cycles2 until

documented tumor

progression or

unacceptable toxicity or

study withdrawal or

death

Obtain whole blood

sample1 Irinotecan pharmacokinetic

blood samples,

D1–3, cycles 1 and 21

Obtain blood samples on Day 21 cycle

2 and at end of treatment visit for

correlative studies1

R

2nd line FOLFIRI +/- regorafenib 1L Anti VEGF/ Anti EGFR / no targeted Tx: 65% / 8% / 28%

PFS (primary endpoint) 6.5 vs 5.3 mo, p=0.0473

OS: 13.8 vs 11.7 mo (NS)

• Positive Phase 2 study

• Confirms the role of

antiangiogenics in the

2nd-line setting


0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12 14 16 18 20 22

HR = 0,72

Log-rank test, p=0,473

Pro

gre

ssio

n f

ree

surv

ival

time (months) No. at risk

A 120 102 81 59 40 22 13 11 7 6 2 1

B 61 47 38 22 14 7 3 3 2

2nd line FOLFIRI +/- regorafenib

O’Neil et al, ESMO 2016

PFS OS

Folfiri Regorafenib 6,5 mo 13,8 mo

Folfiri 5,3 mo P = 0,0473 11,7 mo P = NS

Tabernero et al, Lancet 2015

Folfiri Aflibercept 6,9 mo 13,05 mo

Folfiri 4,67 mo P < 0,001 12,06 mo P = 0,0032

Folfiri Ramucirumab 5,7 mo 13,3 mo

Folfiri 4,5 mo P = 0,0005 11,7 mo P = 0,0219

Bennouna et al, Lancet Oncol 2012

Folfiri Bevacizumab (TML) 5,7 mo 11,2 mo

Folfiri 4,1 mo P < 0,0001 9,8 mo P = 0,0062


2+-line Tx

• FOLFIRI can combine with regorafenib, improves ORR, and is

tolerable with a modified regorafenib regimen

• Sorafenib-irinotecan combo: interesting results but what

development in the future?

• TAS 102 confirms its efficacy and good tolerability in Asia

Samalin et al., Kim et al., O’Neil et al ESMO 2016

LUME: NINDETANIB IN LAST-LINE

Negative for OS but positive for PFS???

PFS: 2.60 vs 1.41 mo

HR 0.57, P<0.0001

Investigator assessment

Van Cutsem et al. ESMO 2016

CCTG CO.23: NAPABUCASIN IN LAST-LINE

Negative for all-comers but positive for Stat3+ patients

Jonker et al. ESMO 2016

L2 Phase III ongoing…

Acknowledgments

Julien Taïeb

Publiclin team

Chiara Cremolini Per Pfeiffer

Florence Huguet Jérémie Lefèvre

Marc Peeters Ian Chau

Heinz Joseph Lenz Pierre Laurent-Puig

Volker Heinemann Claus-Henning Khöne

Eric Van Cutsem Ramon Salazar

the best of esmo 2016...dm has received honoraria for consultancy/advisory boards and symposia from...

Documents